Separating Tectonic, Magmatic, Hydrological, and Landslide Signals in GPS Measurements near Lake Tahoe, Nevada-California Geoffrey Blewitt, Corné Kreemer, William C. Hammond, & Hans-Peter Plag NV Geodetic Laboratory, NV Bureau of Mines & Geology, University of Nevada, Reno with thanks to Ken Smith and David Von Seggern
Separating Tectonic, Magmatic Hydrological and Landslide Signals Two case studies of GPS transients Slide Mountain 2003 Smith et al., 2004, doi:10.1126/science.1101304 Mogul-Sommerset 2008 Both transients coeval with earthquake swarms Swarms of magmatic or tectonic origin? Transients from crustal deformation or landsliding? Transients complicated by hydrological effects
Summary of SLID Transient August 2003 January 2004 Earthquake swarm under Lake Tahoe 29-33 km deep below seismogenic zone Rheological implication rapid stress change Seismic moment Mw 3.8 GPS transient at SLID 3-D displacement ~12 mm Geodetic moment Mw 6.1 Timing is the same as seismic 99% aseismic Evidence magmatic
Depth of Seismicity vs Time (Smith et al., 2004) 1. Quakes are below typical seismogenic depth, which requires a very rapid stress transient 2. Seismic time series starts with upward trend at 2.4 mm / second 8.5 m / hour 1.4 km / week 3. Supports hypothesis of lower crustal magmatic intrusion, migrating upward.
Seismological Analysis (Ken Smith & D. Von Seggern) Deep event has clearly ended on Jan 1 2004 Shallow µ -seismicity follows deep Event Brittle response
SLID GPS GPS Method seasonal and secular removed model during seismic quiescence Fit to transient Blue curve is not GPS!! Cumulative moment from deep earthquakes Curve scaled to fit GPS data N = 4 mm E = 6 mm U = 9 mm Okada model for tensile crack moment Mw 6.1 99% aseismic Earthquakes only a symptom of the process, not the cause East velocity changes! Regional magmatics at work?
SLID Time-Series Filter (pre 2003.5) Global Frame Western US Frame Seasonal Removed
TONO (control) Time-Series Filter (pre 2003.5) Global Frame Western US Frame Seasonal Removed
Mogul Earthquake Swarm, Feb-Aug 2008 (Ken Smith) (Mw 5.0) My house
GPS Site RENO Installed January 2004 Installed January 2004 MAGNET network Now > 300 stations Fortuitous Location Within Mogul swarm 200 meters from Mw5.0 mainshock, April 2008! Smallest earthquake measured by GPS
RENO GPS Co-seismic offset Mw 5.0, 25 April 2008 No pre-seismic signal Large post-seismic slip
Tectonic or Triggered Landslide? The Mogul Landslide Active ~100,000 yr Last slide ~1900 induced by irrigation ditch RENO GPS site lies above the eastern headwall STDI Installed 20 May 2008 Above southern headwall 296 m from RENO STDI RENO
Mogul Landslide Experiment RENO-STDI (296 m) Red = RENO Blue = STDI
Mogul Landslide Experiment RENO-STDI (296 m) Rel. Motion << 1 mm Clearly no landsliding
GPS Regional Strain-Rate Tensor Pre-Seismic MOGL RNO1 VRDE RENO Velocities relative to Sierra Nevada Transtensional Strain = Shear + E-W Extension No-length change aligns with strike-slip seismicity Indicates tectonic response not magmatic
GPS Signals over few km 2 days pre-seismic MOGL southward Consistent with regional inter-seismic strain tensor Strain-Rate Tensor 2 days pre-seismic VRDE westward RNO1 VRDE MOGL RENO RENO northward RNO1 eastward (hydological?)
GPS Signals over few km 2 days pre-seismic MOGL southward Strain-Rate Tensor 2 days pre-seismic VRDE westward RNO1 VRDE MOGL RENO RENO northward Consistent with regional inter-seismic strain tensor RNO1 eastward
Post-Seismic After-Slip Model Estimated logarithmic time constant = 0.9 days
Co- and Post-Seismic Motion are closely aligned
Dislocation Model
Co- and Post-Seismic Slip Model Green: co-seismic (Apr 26) Red: post-seismic (to May 19)
Mogul Swarm Summary Co-seismic displacements of Mw5.0 main shock Measured at 4 GPS stations Landsliding not responsible for detected GPS signal Displacement pattern consistent with regional strain tensor swarm seismicity pattern Post-seismic consistent with after-slip Evidence tectonic source, not magmatic Hydrological signals may confound RNO1 Bad GPS site not attached to bedrock Large seasonal signal estimated and removed Residual signal consistent with tectonics
Velocities Swarm Dilatation Uplift Geothermal Swarm Strain rate Geothermal Swarm Reno-Tahoe Regional Deformation Mix of shear and extension Mogul swarm consistent with high shear localized on Mount Rose fault Incipient fault propagation? Tahoe swarm consistent with magmatics Uplift associated dilitational stress Geothermal area is magmatic?
GPS Vertical Velocity Map Great Valley Huge subsidence Bad stations included Hydrological effects Areas of Uplift Sierra Nevada Mt Lassen, Mammoth CNSB (postseismic) San Andreas Reno-Tahoe Uplift Geothermally active Magmatic evidence
Velocities Strain rate Deformation in RenoTahoe Area Temperature Dilatation Uplift Extensionalmagmatic relationship? (crustal doming)
Separating Signals: Lessons Learned 1. GPS+micro-seismology is powerful combination Pattern + timing Use all evidence to construct displacement model 2. Exploit data during relative seismic quiescence: GPS Velocities Regional Tectonics Process(t,x)? GPS Seasonal Signals Hydrological Loading Identify hydrological effects @ bad GPS sites 3. Consider greater regional context What is typical or unusual? 4. These techniques identify magmatic-tectonic interaction in the Reno-Tahoe area